EAS Time Structures with ARGO-YBJ experiment

1 - INFN-CNAF, Bologna, Italy2 - Università del Salento and INFN Lecce, Italy

A.K Calabrese MelcarneA.K Calabrese Melcarne11, , G.MarsellaG.Marsella22 for the ARGO-YBJ Collaborationfor the ARGO-YBJ Collaboration

XXIII European Cosmic Ray Symposium, XXIII European Cosmic Ray Symposium, ECRS 2012 Moscow, Russia, July 3-7, 2012ECRS 2012 Moscow, Russia, July 3-7, 2012

1

High Altitude Cosmic Ray Laboratory @ YangBaJing,Tibet, ChinaSite Altitude: 4,300 m a.s.l. , ~ 600 g/cm2

The ARGO-YBJ experimentThe ARGO-YBJ experiment

ARGO-YBJ

2

Strip = space pixelPad = time pixelTime resolution ~1.8 ns

10 Pads (56 x 62 cm2)for each RPC

8 Strips (6.5 x 62 cm2) for each Pad

78 m111 m

99 m

74 m

(43 m2)1 CLUSTER = 12 RPC

RPC

analog charge read-out dynamical range up to ~ 104 TeV

3

Num

ber o

f Fire

d St

rips

Full coverage, high time and space resolution provide a

detailed view of shower front

Data AnalysisData Analysis

- Quality cut on S2

Reconstruction

Time sequence and position of hit pads used to reconstruct the CR

arrival direction and core position

- Using a plane (α=0)

- Using a conical correction (α0)

- Core reconstructed within the central carpet

contamination of mis-reconstructed events less than 10% at low multiplicity, rapidly decreasing at higher multiplicity

4

Average Curvature:Average Curvature: the mean of time residuals Δt(R) with respect to a

plane fit

Time profileTime profile

- larger than 10 ns for particles landing further than 60 m from the core.

- no significant dependence on pad multiplicity observed.

7.9 108 events

zenith < 15°

5

Data vs SimulationsData vs Simulations

Very good agreement at the level of time profile

curvaturecurvature Shower generatorShower generator Corsika 6.720 with

SIBYLL+FLUKA as hadronic interaction models at high and low energies

~ 107 proton showers dN/dE E- (=1) 300 GeV - 100 TeV

Detector simulationDetector simulation GEANT3

zenith < 15°

6

Large RMS Shower frontsLarge RMS Shower fronts

Wide showers

Double Front showers

• In order to exploit at maximum space-time information, we started a detailed study on the longitudinal time structures in data

• The idea is to study more in detail the shower structures in order to define selection criteria for particular analysis (gamma/hadron separation, composition, exotic physics)

• In particular we studied showers with large time residual with respect to the shower front

• Two Categories have been observed

7

Double Shower FrontsDouble Shower FrontsReconstruction

– Separation of subshowers– Planar Fit on subshowers– Quality cuts on reconstructed

subshowers • S2 < 100 ns2

New Selection• Hit Number >100 (10% of Ev.)• Fit on time distribution

• Used TSpectrum class of root

Distance of two peaks > σ1+σ2

8

Double Shower FrontsDouble Shower Fronts• What are these showers?accidental coincidences and possible delayed showers

• How many events?

• How many are compatible with accidental coincidences?

– Angular difference distribution– Nhit Distribution– Time delay– Observed vs Expected events

9

Angular difference Angular difference DistributionDistribution

10

Angular differencebetween the reconstructedsubshower directions

In agreement with CR distribution of consecutiveevents

Multiplicity DistributionMultiplicity Distribution

11

Multiplicity distribution of the reconstructedsubshowers Shower 1 is the one whichtriggers the detector!We expect lower multiplicity distribution from shower 2 (triggernot required)

Time DistanceTime Distance

12

How to define time distance betweentwo subshowers?

At least 2 variables:•Peak mean value•T0 from the shower plane Fit

Time DistanceTime Distance

13

Both variables can be used as indicator

Reconstucted AnglesReconstucted Angles

14

Quality requirements on subshowersreconstruction are applied (S2<100ns2)

Applying a planar fit, the directionof both subshowers is reconstructed

AnalysisAnalysisObserved rates

– 7.9 x108 events have been processed

– 700 x103 events selected as double coincidences

– 370 x103 events selected as double coincidences with quality cuts on subshowers

Expected rates• DAQ rate = 3.5±0.1 kHz (multiplicity

of Events Nhit > 20 in a time window τ = 2 μs)

• Rate of observed showers with S2 < 100 ns2 :

λ1= 2.7±0.1 kHz• Expected double coincidences:

λexp = 2 λ1 λ1 τ η = η (28±1) Hz (η is the event selection efficiency)

η x 0.8% of the events

4.5x10-2 % of the events

(1 order of magnitude less than expected if η=1 )

15

AnalysisAnalysisSimulated double events• Artificial double Events have been generated from real data• Merged two consecutive events shifting the time of each hit of the

second event by a randomly extracted ΔT compatible with the 2μs trigger window

• Verified the random double shower distributions (Angle, Multiplicity, relative time distribution)

• Tested the double shower selection and reconstruction efficiency η : The efficiency η is 5% (preliminary selection algorithm) This explains the difference between expected and observed events expected :η x 0.8% = 4.0x10-2 % of the events

observed :4.5x10-2 % of the events

16

In this work the shower time profile and distribution have been presented. The curvature of the showers around the core have been studied, illustrating the reconstruction effect on ARGO-YBJ data.Longitudinal time structures have also been investigated selecting events with large time distribution around the shower front. The attention has been focused on double front showers and , in particular, to identify the expected random double coincidences.A new algorithm have been introduced, increasing the detection efficiency. More efforts are on the way. The detailed study of the angular and time distribution of the selected events will be useful in studying possible shower anomalies. In particular narrow Angular difference events will be analyzed on large statistical sample in order detect anomalies in spatial, time and multiplicity distributions as a flag of possible ‘exotic’ physic events in CR.

Conclusions Conclusions

17

This is a preliminary step to better investigate the possible “physics of multiple shower fronts”